Adaptive Disturbance Rejection Control using System Input -Output Data

A stable discrete -time adaptive disturbance rejection control law is designed for linear systems . The disturbance signal is assumed to be unknown , but can be described by a sum of periodic signals . The baseline control law is based on dynamic inversion with pro portio nal -integral augmentation. D ynamic inversion is carried out using a system input out put model that implicitly contains the effects of the external disturbance . The input output model is realized in an off -line manner using the available input -output data. On -line adaptation of this control law is carried out by providing a parameterized a ugmentation signal to the dynamic inversion. The parameters of this augmentation sig nal are updated to achieve the nominal desired error dynamics . Contrary to the Lya punov -based approaches that guarantee only stability, the current approach guarantees sta bility as well as performance. Simulation results are provided to illustrate the overall control approach.